Building fintech in Kenya — constraints that shape the product

If you import a US or European fintech playbook into Kenya unchanged, you will ship a beautiful sandbox. Production will teach you the rest.

We have built lending and payment-adjacent products here long enough that a few constraints show up every time. They are not blockers — they are the brief.

M-Pesa is not a checkout plugin

In many markets “payments” means a card SDK and a webhook. In Kenya, money often moves on M-Pesa: STK Push for collection, paybill and till for business inflows, B2C for disbursements.

That changes product design. Approval without a reliable payout path is incomplete. A borrower app that lies about disbursement status is worse than a slower app that tells the truth. We wrote more on this in Shipping M-Pesa in production and shipped it end-to-end on Okolea.

Operators are users too

Credit officers, school bursars, and support teams live in the admin portal. If reconciliation means exporting three CSVs and guessing, they will invent a spreadsheet sidecar — and your “system of record” stops being one.

We sit with operators early: who retries a failed payout, who can override a decision, what “pending” must mean on a busy Monday. That conversation shapes state machines more than any framework debate.

Connectivity and devices are part of scope

Assume intermittent networks and a wide device range. Offline-friendly sync, clear retry UX, and APIs that tolerate duplicate callbacks are product features — not nice-to-haves for a later phase.

Audit trails beat slide-deck compliance

Fintech buyers (and regulators) ask who approved, who disbursed, and what the gateway returned. Design for that from day one: immutable event history, role-aware actions, and exports ops can defend.

What we tell buyers on the scope call

1. Name the money path before the UI polish.
2. Budget time for failure modes, not only happy-path demos.
3. Keep senior owners available through first production weeks — payment cutovers are where black-box handoffs fail.

If you are mapping a lending, fees, or payouts product for Kenyan users, book a scope call. We will stress-test feasibility against these constraints — including if another partner or a simpler SaaS path fits better.